The Clinical Risk Management in a Hospital Ward:
a Case-Study adopting System Dynamics Approach
Aiello Valentina
Department of European Studies and International Integration
University of Palermo (Italy)
Francesco Ceresia
Department of European Studies and International Integration
University of Palermo (Italy)
Amedeo Claudio Casiglia
Risk management manager
ASP6 Palermo (Italy)
1. Introduction
In recent years, the attention to medical errors increasing due to a higher patients awareness
about their rights and the importance attributed by health care companies to the quality of their
services. In the context this growing importance it became essential to define patient safety. Even
though the Institute of Medicine IOM defined patient safety as “the avoidance, prevention and
amelioration of adverse outcomes or injuries stemming from the process of healthcare” there is not
yet a full acceptance of this definition by all healthcare organizations because of the strong
distinction, implemented in healthcare organizations, between quality and safety. To overcome this
barrier it is useful to adopt the patient safety definition provided by Emanuel and Berwick’: “patient
safety is a discipline in the healthcare sector that applies safety science methods toward the goal
of achieving a trustworthy system of healthcare delivery. Patient safety is also an attribute of
healthcare systems; it minimizes the incidence and impact of, and maximizes recovery from,
adverse events.”
In Italy, as happened in the rest of the world, the attention to the problem of error in medicine
shown a renewed interest after the publication of the report “To Err Is Human” by the Institute of
Medicine (Kohn et al. 1999). After this important document, the scientific research in the field
started to expand and new frontiers in the study of clinical errors have been opened. The change
of vision generated by this research has “humanized” the professionals working in the health care
system, and has changed the concept of medical errors being no longer considered as a reason
for punishing the guilty yet as a reason for learning from the errors at the level of the whole
organization.
The clinical risk refers to the probability that a patient suffers damage during the delivery of
healthcare services. In order to allow a clear understanding of the phenomenon, it is useful to
define some Clinical risk management (CRM) key-concepts on the base of the indication provided
by the J oint Commission’:
2K. Henriksen, et al., Advances in Patient Safety: New Directions and Alternative Approaches (Vol. 1: Assessment).
Rockville (MD): Agency for Healthcare Research and Quality, 2008.
2 The Joint Commission is an independent not-for-profit organization that accredits and certifies more than 20,000
healthcare organizations and programs in the United States. It is recognized nationwide as a symbol of quality that
reflects an organization's commitment to meeting certain performance standards. The stated mission is: "To continuously
improve healthcare for the public, in collaboration with other by evaluating healthcare organizations and
inspiring them to excel in providing safe and effective care of the highest quality and value". J oint Commission
International is also one of the most it ization providing i i healthcare accreditation services to
hospitals around the world.
e sentinel event is any unanticipated event in a healthcare setting resulting in death or
serious physical or psychological injury to a patient, not related to the natural course of the
patient's illness. Such events are called “sentinel” because they signal the need for
immediate investigation and response. The terms “sentinel event” and “error” are not
synonymous; not all sentinel events occur because of an error, and not all errors result in
sentinel events. Sentinel events are 16 and they were been defined by the Joint
Commission (table 1).
e adverse event is any untoward medical occurrence in a patient or clinical investigation
subject administered a pharmaceutical product and which does not necessarily have to
have a causal relationship with this treatment. An adverse event (AE) can therefore be any
unfavorable and unintended sign (including an abnormal laboratory finding, for example),
symptom, or disease temporally associated with the use of a medicinal product, whether or
not considered related to the medicinal product.
e near miss is a situation which could cause an adverse event for the patient (for example, a
fall avoided by the intervention of a nurse).
e no harm event is an event that had the potential to result in harm to the patient (such as
falling without patient outcomes).
Different types of sentinel events
Procedure on wrong patient
Procedure on wrong site
Wrong procedure in correct patient
Unintended retention of foreign object in a patient
Hemolytic blood transfusion reaction resulting from ABO incompatibility
Medication error leading to the death of patient reasonably believed to be due to
incorrect drugs administration
7 | Maternal death or serious morbidity associated with labor or delivery
8 | Death or permanent disability in healthy newborn weight more than 2500g
9 | Death or serious injury as a result of patient fall
10 | Suicide or attempted suicide of a patient
11 | Rape or assault of patient
12 | Rape or assault of staff member
13 | Death or serious injury as a result of intra-and extra-hospital carriage
14 | Death or serious injury due to improper triage assignment
15 | Death or serious injury as a result of surgery
16 | Any other adverse event that cause death or serious injury
| a] Ww) ]
Table 1. Sentinel events
2. Literature Review
Thirty years ago medical errors weren't even mentioned in medical literature. In 1994 Leape
published a paper about the question of medical errors®. In this work, the author underlined that
error rates in medicine were particularly high and that the healthcare companies had not take into
account it properly, as other safety critical industries had. He argued that the solution to the
medical errors’ problems did not lie in medicine but in different field as a psychology and human
factors. Many errors are beyond the human’s conscious control so error prevention that relies
3 Leape LL. (1994). Error in medicine. J ournal of the American Medical Association, 272(23):1851-7
exclusively on discipline and training is doomed to failure. For this reason he focused his attention
on the way to change or improve the work condition than the personnel training.
In the year 2000‘ Reason re-published a paper of 1990° about error management, which
addressed the medical error’s issue from a new perspective. Following the theoretical framework of
Reason (2000), the human error problem can be viewed in two ways: the person approach and the
system approach. Each model of error refers to different views of errors etiology and
management. The person approach focuses on the unsafe acts (errors and procedural violations)
of people at the sharp end, arising from deviant mental processes (forgetfulness, inattention, poor
motivation etc.). Their management is aimed at reducing undesired variability in human behavior.
According to Reason, “followers of this approach tend to treat errors as moral issues, assuming
that bad things happen to bad people” (Reason, 2000). The system approach refers to the concept
that errors can occur even in the best organizations because of the fallibility of people. The errors
management is based on the assumption that although “we cannot change the human condition,
we can change the conditions under which humans work” (Reason, 2000).
A'SWISS CHEESE’ RISKLINE UP
Latent conditions
poor design, procedures,
Patient safety incident
management decistons etc
Active errors
pentane sy
Figure 1. The “Swiss cheese” model, Reason The human error, 2000.
A central role, in the system approach, is occupied by defenses and barriers of the organization. In
an ideal and desired condition, each defensive barrier would be intact, however, in real life, they
were more likely to be slices of “Swiss cheese” (figure 1) with many holes constantly moving,
opening and shutting. As Reason said: “the presence of holes in any one “slice” does not normally
cause a bad outcome. Usually, this can happen only when the holes in many layers momentarily
line up to permit a trajectory of accident opportunity bringing hazards into damaging contact with
victims” (Reason, 2000). The Swiss cheese model represents a metaphor of the trajectory of an
accident which gives us the sense of hazard being ever present and occasionally breaking through
when all the holes in the Swiss Cheese line up. In the light of these studies, a healthcare
organization can be considered as a complex system, inside which it is possible to trace the
various interconnected sub-systems.
4 Reason, J . (2000). Human error: models and management BMJ , 320, 768-770.
Reason, J ames (1990): Human Error. New York, NY, Cambridge University Press.
Organic caarlbutin Care Delivery
and Culture a “ Problems
Factors
Defencess
No]
|__Jifonment | vasa sets
‘Management
Decisions
and
|_| Team Factors |__|
Errors
Processes | 4] Task Factors
}-—>} Patient Factors |__.!
LATENT
CONDITIONS
Figure 2. Organizational accident model (Reason 2000)
Through the work of Reason, the evolution of error has gone from an individual perspective to a
system view. In the individual perspective, the efforts to reduce the errors are centered on people
and are based on the encouragement to "do better" (upgrading or adding new rules or
procedures). The prevailing culture is the “blame culture”. In the systemic perspective, errors and
human behavior cannot be understood in isolation, but only in relation to the context in which
people work. Many of the errors need to be considered from this broad system approach to be
managed and Reason provides an efficient organizational accident model (figure 2) to describe the
immediate errors and problems and the background latent conditions.
Vincent (1998), by using the system approach provided by Reason (1990), has studied the role of
human factors in the generation of an adverse outcome in healthcare. The term human factors can
be defined in several ways but, a widely accepted definition is that of the Health and Safety
Executive: “Human factors refer to environmental, organizational and job factors, and human and
individual characteristics which influence behavior at work in a way which can affect health and
safety. A simple way to view human factors is to think about three aspects: the job, the individual
and the organization and how they impact on people’s health and safety-related behavior.” (HSE,
1999).
In reference to the human factors that can influence clinical practice and can contribute to the
generation of the adverse outcome, Vincent (1998) identifies 7 main framework and their relative
contributory factors (table 2).
Framework Contributory Factors E ples of Probl That C i to
Errors
Institutional Regulatory context Insufficient priority given by regulators to safety
Medico-legal environment issues;
National Health Service Executive Legal pressures against open discussion,
preventing the opportunity to learn from adverse
events
Organization | Financial resources and constraints Lack of awareness of safety issues on the part of
and Policy standards and goals senior management;
management | Safety culture and priorities Policies leading to inadequate staffing levels
Work Staffing levels and mix of skills Heavy workloads, leading to fatigue;
environment | Patterns in workload and shift Limited access to essential equipment;
Design, availability, and maintenance | Inadequate administrative support, leading to
of equipment reduced time with patients
Administrative and managerial support
Team Verbal communication Poor supervision of junior staff;
Written communication Poor communication among different professions;
Supervision and willingness to seek | Unwillingness of junior staff to seek assistance
help
Team leadership
Individual Knowledge and skills Lack of knowledge or experience;
staff Motivation and attitude Long-term fatigue and stress
Member Physical and mental health
Task Availability and use of protocols Unavailability of test results or delay in obtaining
Availability and accuracy of test results | them;
Lack of clear protocols and guidelines
Patient Complexity and seriousness _ of | Distress;
condition Language barriers between patients and
Language and communication caregivers
Personality and social factors
Table 2. Framework of Factors Influencing Clinical Practice and Contributing to Adverse Events (Vincent et al., 1998)
In this hierarchy of factors, patients and staff as individuals are at the front-end (bottom) of the
factors, team factors and working conditions in the middle, and organizational/institutional factors
at the top.
e Patient framework: the patient’s condition has the most direct influence on practice and
outcome, moreover patients have a key role to play in helping to reach an accurate
diagnosis, in deciding about appropriate treatment, in choosing an experienced and safe
provider, in ensuring that treatment is appropriately administered, monitored and adhered
to, and in identifying adverse events and taking appropriate action. Other patient factors
that may influence the communication with the staff and hence the clinical practice are
patient's language and personality.
e Task framework: this framework refers to the procedures to implement in order to ensure
the correct dispensing of healthcare. We have to take into account the availability and use
of protocols and their lack of clarity as key factors to affect the quality of care and in the
generation of adverse outcome.
e Individual staff member framework: several individual factors may influence the clinical
practice such as personality, knowledge, experience, and training. Other factors related to
staff concern the physical and mental health. These last factors may be damaged by work
stress and burn out, two problems strongly associated with healthcare contexts.
e Team framework: each member of the staff is part of a team, and his performance may be
influenced by other members, and how they are organized, and how they support,
supervise, monitor, and communicate with each other.
e Work environment framework: factors related to the work environment may include staffing
levels and mix of skills, patterns in workload and shift, design, availability, and maintenance
of equipment. There is a growing evidence base from healthcare architecture, interior
design, and environmental and human factors engineering that supports the assertion that
safety and quality of care can be designed into the physical construction of facilities.
e Organizational and management framework: organizational factors are omnipresent, but
difficult to quantify (organizational climate, group norms, morale, authority gradients, local
practices) that often go unrecognized by individuals because they are so deeply immersed
in them. However, over time these factors are sure to have their impact and the system
performance. Management factors refer to the way in which the organization is handled
such as financial resources and constraints and policy standards and goals.
e Institutional framework: the institutional framework influences patient safety and quality of
care by shaping the context in which care is provided. It refers to the economic and
regulatory context, and to the National Health Service Executive.
The seven levels framework has outlined the patient, task and technology, staff, team, working
environment, organizational and institutional environmental factors that are revealed in analyses of
incidents. These same factors also point to the means of intervention and to the different levels on
which safety and quality must be addressed.
3. Traditional CRM’s method and their pitfalls
CRM is an approach for improving the quality and safe delivery of healthcare. This can be
accomplished by placing special emphasis on identifying conditions that put patients at risk, and by
establishing mechanisms to minimize or prevent these risks.
In order to reduce the incidence of the clinical risk and to improve the quality of the cares, in the
healthcare sector have been imported risk management methods successfully applied in other
sectors. The most widespread methods are:
a) Root Cause Analysis,
b) Clinical Audit,
c) Incident Reporting.
All these methods, although useful in the identification of risk probability and assessing the
potential effects of the occurrence of adverse events, do not support the management of
healthcare organizations in the identification and evaluation of risk management policies because
they are based on an linear analysis of the system and the relationships between business
processes. In particular, the shortcomings of those methods are:
e They don’t take into account the feedback structure of the net of causality that links each
other the variables of the healthcare system;
e They are static, namely they ignore delays normally existing between the triggering of the
cause and the occurrence of the related error and, consequently, they are not suitable to
simulate future trends;
e They don’t consider the interactions between the different risks;
e They are inadequate in helping healthcare companies in setting safety targets and
evaluating safety performance improvement on a quantitative basis;
e They don’t properly support the healthcare companies’ management in the identification
and assessment of policies aimed at improving the clinical risk profile.
e They don’t take into consideration the costs and their effect on the organizations
management.
For these reasons, all these methods are inadequate in helping healthcare companies in setting
safety targets and evaluating safety performance improvements. Therefore, it is possible that a
hospital does not invest in clinical risk reduction because of the costs and of “heaviness” of the
operational procedures in such investment. This happens because of the lack of understanding
and/or of ‘inability to assess the benefits that investments in clinical risk reduction produce. In fact,
where these methods have been considered often they have not had a real application, because
healthcare organizations have been limited to a formal implementation of these procedures, but
this has resulted in a substantial improvement in the approach of clinical risk management.
Therefore, it is necessary that healthcare companies perceive that an improvement in the risk
profile often results in a considerable saving on insurance policies, on the cost for claims and on
the costs of “non-quality” and “safety”. In addition, by reducing the clinical risk, the healthcare
companies get a better image and, therefore, an increase in their competitiveness. Therefore, it is
necessary to adopt a systemic and multi-dimensional approach that allows healthcare companies’
managements to properly evaluate the effects of CRM policies on organizations’ performance, both
in the short and medium long term.
4, System dynamics as a new powerful method to deal with the clinical risk
Despite the relative newness of the adoption of the SD methodology in the CRM field different
examples of applications of the system dynamics approach to the healthcare sector have been
reported in the literature. The system dynamics literature on this topic can be classified into two
groups: those that deal with specific diseases and those that deal with broader policy and
management concerns. Literature focusing on diseases includes: Oral Health (Hirsch et al., 1975);
Cardiovascular Disease (Hirsch and Myers,1975 & Luginbuhl et al., 1981); Diabetes (Homer et al.,
2004 & Jones et al., 2006); Obesity (Homer et al., 2006); and chronic illnesses more generally
(Hirsch and Immediato, 1999). Literature focusing on management includes: EHIR Adoption (Erdil
& Emerson, 2008); Patient flow (Wolstenholme, 1999); Safe Design Capacity (Wolstenholme et al,
2007); and Waiting Lists (Van Ackere and Smith 1999)°.
According to Homer and Hirsch “a central tenet of system dynamics is that the complex behaviors
of organizational and social systems are the result of ongoing accumulations (of people, material or
financial assets, information, or even biological or psychological states) and both balancing and
reinforcing feedback mechanisms” (Homer and Hirsch 2006). System dynamics uniquely offers the
practical application of these concepts in the form of computerized models in which alternative
policies and scenarios can be tested in a systematic way that answer both "what if” and “why”.
Such models support the healthcare management in properly evaluating the effects of CRM
policies on organization performance, both in short and medium-long term.
5. Case-study: the ward of obstetrics and gynecology in Italian Hospital
The research was carried out in a public hospital placed in a little town near to Palermo (the capital
of Sicily Region), which serves approximately a population of 35.000 people (some macro-
variables of the hospital and the operational unit involved in the research are shown in the table3).
The hospital is part of the ASP6 Palermo that represents the Provincial body for Health Services,
an entity as provided by law that is responsible to manage and coordinate the services and public
health activities for the whole province. In particular it was decided to concentrate the research ina
specific operational unit: the ward of obstetrics and gynecology.
®Goldsmith D., Siegel M. (2011) Improving Health Care Management Through the Use of Dynamic Simulation Modeling
and Health Information Systems. System Dynamics conferences 2011 proceed papers.
MACROVARIABLES HOSPITAL OBSTETRICS AND
GYNECOLOGY WARD
BEDS 87 ORDINARY 12 ORDINARY
18 DAY HOSPITAL 2 DAY HOSPITAL
EMPLOYEES 291 (85 DOCTORS, 154 NURSES, 33 (10 DOCTORS, 10
2105S, 2 OTA, 8 AUXILIARY, 21 OBSTETRICIANS, 12
CS) NURSES, 10SS)
AVERAGE PATIENTS PER YEAR 4124 IN 2011 720 IN 2011
4238 IN 2012 914 IN 2012
4025 IN 2013 773 IN 2013
RISK MANAGER YES
SYSTEM OF COMPLAINS YES YES
MANAGEMENT
INCIDENT REPORT 32 (2011) 16 (2011)
29 (2012) 1 (2012)
14 (2013) 1 (2013)
CLAIMS 11 (2011) 2 (2011)
37 (2012) 4 (2012)
23 (2013) 2 (2013)
SENTINEL EVENTS 1 (2011) 0 (2011)
1 (2012) 1 (2012)
0 (2013) 0 (2013)
Table3. Hospital and obstetrics and gynecology ward macro-variables
The choice is motivated by the high risk level of the operational unit at issue, and by the high
emotional involvement that the ward of obstetrics and gynecology involves, activating social,
ethical and emotional values, extremely important as they concern motherhood and the newborn.
Moreover it was decided to focus just on the obstetric and gynecology ward and not on its
outpatient clinic. The reason is primary the following: the possibility of errors occurrence in a
outpatient clinic (sentinel, near miss or no harm event) is low because of the service characteristics
(there are no drug administration, no surgery, no blood transfusions in a outpatient clinic and so
on).
5.1 Objective of the study
The objective of the research is to investigate the relationship between the patient safety culture in
the hospital’s ward and various types of adverse outcome in the clinical practice. The hypothesis is
that a safety culture assessment represents a tool for improving patient safety. While a variety of
levers (clinical training and guidelines, information technology, organizational structures and
industry regulations) are being pushed in healthcare organizations to improve patient safety, the
belief is growing that an institution’s ability to avoid harm will be realized only when it is able to
create a culture of safety among its staff. Safety culture is a performance shaping factor that
guides the many discretionary behaviors of healthcare professionals toward viewing patient safety
as one of their highest priorities’. In order to transform culture it is important to first understand and
confront it, to do this a questionnaire was administered to the whole ward’s population. The choice
questionnaire is the “Hospital Survey on Patient Safety Culture” (Nieva, Sorra, 2003).
By using a simulation model we want to test the weight that the patient safety culture has in the
generation of adverse events in order to identify the best policy able to achieve the optimal trade-
off between cost and benefit.
T Nieva V.F., Sorra J. (2003) Safety culture assessment: a tool for improving patient safety in healthcare organizations.
Qual Saf Health Care2003;12(S uppl II):ii17-ii23
5.2 Research methodology
5.2.1 Exploring the CRM procedures in the hospital
In order to explore the CRM procedures in the hospital three semi-structured interviews were
administered to the main hospital’s key actors of risk management: the ASP risk manager, the
operational unit referent for clinical risk, and the responsible of midwife’s nursing care.
The hospital is part of the ASP6 Palermo that represents the Provincial body for Health Services,
an entity as provided by law that is responsible to manage and coordinate the services and public
health activities for the whole province. Inside ASP6 is planned a risk management operational unit
aims to organize the hospitals clinical risk management according to the guidelines, it has
implemented a system clinical risk management which requires the involvement of Hospitals and
individual wards (1 ASP risk manager, 5 risk management medical directors one for each ASP
hospital, 41 health workers one for each ward of the five ASP hospitals).
Over the past years three different training courses on the clinical risk management have been
performed. The first training course was addressed to the 41 operational unit referents for clinical
risk, the second was addressed to nurses. Both training courses focused about the diffusion of a
patient safety culture in order to improve the staff participation in the clinical risk management. The
third training course was addressed to nurses, it was about the management procedures and
transmission of incident reporting and prevention of patient's falls.
The Ministry of Labor, Health and Social Policy in 2010 has set up a data stream, the Information
System for Monitoring Errors in Healthcare SIMES, with the aim of detect information relating to
sentinel events and claims. SIMES data stream it is composed of three different forms to be filled:
form A, form B and form 3 claims. The health care worker involved in the adverse event, fills and
submits the form A with the operational unit referents for clinical risk to the risk management
medical directors of the hospital within three days. The risk management medical directors with the
ASP risk manager immediately starts a preliminary analysis to determine if the event meets the
criteria to be called a sentinel event. In the case where the event is defined sentinel event he
transmits the data to the Ministry of Health. Within seven days from the date of the event is set up
an internal committee shall carry out a Root Causes Analysis and defines any improvement
measures. The risk management medical directors and the ASP risk manager in the light of the
results of the committee fill out and submit, no later than forty-five days from the date of the event,
the form B to the Ministry of Health.
In addition to the SIMES data stream other sources of data collecting were activated in the ASP6:
spontaneous reporting of adverse events, near miss and no harm event (Incident Report),
monitoring implementation of the Joint Commission International standards and Ministry
recommendations. These streams are inserted in a database. However the Incident Reporting, just
because spontaneous, it is not always able to keep track of all near misses and no harm events
that occur in the organization. With regard to the third category of events reported by the incident
report namely adverse events, they are usually associated with claims because they are related to
events occurred and not to not happened events (near miss) or events with no harm for the patient
(no harm events).
However this way to manage the clinical risk at ASP6 is quite new, it was introduced in 2011, for
this reasons we have no data before 2011 and moreover, for this reason we have not a shared
patient safety culture among the hospital staff, the culture needs time to become part of the work
routine. Due to this recent CRM practice adoption we have no a complete track of error occurred,
as we can see from hospital data just few events were reported to the management during this last
three years (see table3).
5.2.2 Exploring the professional staff's CRM culture in the hospital
A questionnaire was administer to the whole personnel of the ward (10 doctors, 10 obstetricians,
12 nurses and 1 OSS), in order to explore the perception about the patient safety culture and
clinical risk in the hospital. Responses were voluntary and no personal information was collected to
avoid fear of respondents’ identification. Twenty-seven of the thirty-three questionnaires
administered were completed and returned. The chosen questionnaire is the “Hospital Survey on
Patient Safety Culture” Nieva, Sorra, 2003 (see annex 2). The questionnaire is focused on patient
safety issue and on error and event reporting. The dimensions investigated by the questionnaire
are:
Seven unit-level aspects of safety culture:
e Supervisor/Manager Expectations & Actions Promoting Safety (4 items),
e Organizational Learning—C ontinuous Improvement (3 items),
e Teamwork Within Units (4 items),
e¢ Communication Openness (3 items),
e Feedback and Communication About Error (3 items),
e Non-punitive Response to Error (3 items), and
e Staffing (4 items).
Three hospital-level aspects of safety culture:
e Hospital Management Support for Patient Safety (3 items),
e¢ Teamwork Across Hospital Units (4 items), and
e Hospital Handoffs and Transitions (4 items).
Four outcome variables:
e Overall Perceptions of Safety (4 items),
e Frequency of Event Reporting (3 items),
e Patient Safety Grade (of the Hospital Unit) (1 item), and
e Number of Events Reported (1 item).
The Hospital Survey on Patient Safety Culture (Hospital SOPS) was originally developed, pilot-
tested and revised in the USA and then released by the Agency of Healthcare Research and
Quality (AHRQ). The survey was designed to assess opinions of hospital staff about patient safety
issues, medical error and event reporting and includes 42 items measuring the above mentioned
12 dimensions of patient safety culture. Respondents are asked to rate each item of a dimension
on a five-point likert scale of agreement (strongly disagree, disagree, neutral, agree and strongly
agree) or frequency (never, rarely, sometimes, most of the time, always). The survey includes two
questions asking respondents to provide an overall grade on patient safety for their work area/unit
and to indicate the number of events they have reported over the past 12 months. Respondents
are asked to provide limited background information about themselves.
5.2.3 Designing the causal loop diagram (CLD) and built the Stock and Flow
Model
A section of group model building GMB with the three main hospital’s key actors of risk
management was used to collect data for the Causal Loop Diagram CLD. The GMB represents the
first step in the system dynamics modeling process, allowing to create a shared view of the
problem among the key stakeholders. Following Vennix et al. (1992), three main tasks were
performed by modelers to generate the CLD: elicitation of information, exploring courses of action
or convergent tasks, and evaluation. The main output of the GMB sessions is the CLD, a document
that describes the causal relationship between the key-variables of the healthcare company
involved in this study, in order to understand their role in the etiology of the adverse event.
The CLD was the starting point to build a Stock and Flow Model, in order to test the impact of
different policies on ward's performance. The model was built using Powersim Studio 7.0.
6. Results
6.1C ies from q i ire
The data collected in the questionnaire allowed to identify the perception of the personnel about
the issue of clinical risk management. The data analysis was performed by SPSS 16.0, and it was
analyzed the correlation (r di Pearson) between the different dimensions (annex! correlations).
Moreover the collected data have allowed to determine the overall patient safety index (66,54%) by
calculating the scale means.
6.2 The causal loop diagram
The data collected with the group model building session allowed to identify some of the main
cause-effect relationships characterizing the organizations system. As showed in the causal loop
diagram (figure 3), the high number of patients is related to a high number of treatments. The law
of large numbers tells us that a higher number of treatments determine a rise in the number of
adverse events due to clinical errors. The high number of adverse events cause an increase in
claims that are related to a loss of image. The loss of image has a return in term of loss of patients
(loopB1).
total costs
+
treatments earnings
R4 +
insurance costs
, financial availabilty
—_— is
52
+
hospital image CRM policies
Ov ane: ome J
staff skills
(communication,
ee, empowerment)
A aiverse event,
R3
+
structural improvement
(more rooms Ta
Figure 3. The Causal Loop Diagram
Nevertheless, when the number of treatments increases, there is an increase in treatments
earnings and consequently in the financial availability. In this way the health care company has
more money to invest in CRM policies. It is logical that the CRM policies reduce the financial
availability (loopB2). However CRM policies should improve some staff skills (such as
communication), or they should activate some structural improvement (more rooms and
equipments, computerized medical records), in order to reduce the number of adverse events and
the number of claims, with a positive effect on the hospital image and consequently in the number
of patients and treatments (loopR2 and loopR3). A reduction in the number of adverse events
causes a reduction of claims. In this way the costs related to insurance and compensations claims
and hence the total costs are reduced. This reduction has a positive effect on the financial
availability and so in the possibility to invest more in CRM policies with a further reduction of
adverse events (loopR4).
6.3 The Stock and Flow Model
Based on the CLD described above a stock and flow structure has been developed in order to test
the impact of different policies on ward’s performance. Figure 4 shows a section of the stock and
flow model describing the hospitalization process: subsystem1 patient.
sjuagedul
Jeydsoy sayjo
abeiane =" seen
ayes Anuedna30 paq quanedul ssanaugoemy 0) awn abesany
Aypeded yuayed ‘woraq 0} aw | leudsoH S CY
ri —“O)
yawn aBesany aye quayedui aye
‘ayer AQuedno320 paq sjeydsoy ayo juaugeay juajed
ane) stata leydsoy sayjo
quaujeeg quaned ayer quaqedut quauqeay, ay uoneindod — jes\6oj038uK6 Aq
paGseysip quaged Suge By “105 Bunem uogeindod ~~*--~ payaye uoneindod
a Om
xapuj Wopue Yy oO aGseyrsi
ayes Aanored posseysip
oe ayer £19028) HO quanedino:
ssad0.d uogezi|edsoy auy 0} payejau aunqon.gs Moy pue y20}s *TWAISASUNS “p aun614
The central stock “population affected by gynecological event” represents the amount of people,
served by the hospital and affected by gynecological events that require hospital treatments. As a
consequence these people can become hospital inpatient or became other hospital inpatient due
to the “hospital attractiveness”. For this reason, from the stock “population affected by
gynecological event” we have to possible outflow: one is “other hospital inpatient rate” that conduct
people to be treated in other hospital and then discharged and placed again in the main stock
“population affected by gynecological event’; and the other one is “population waiting for treatment
rate”.
This second flow is the beginning of patient chain: a patient passed from several stock and flow
(‘population waiting for treatment”, “inpatients rate”, “inpatients”, “patient treatment rate”, “patient
discharged” and “recovery rate”) that represent the hospitalization and the recovery process. This
process is affected by two indexes: “patient capacity” (number ward’s beds), and average time to
treat (a means of different treatment timing of the ward over time).
rAT
LC) no harm event >
FAT % of no harm
events
% of no harm no harm event rate
events
Wa
patient treatment
medical error )
near miss rate,
ate poten inedlcal
LL ee 0
- potential medical
effect of patient error Tate
safety index on OC /
ts
% of adverse
MK
% of sentinel event events
% ofnearmiss correct treatment
ro Time to define from near miss
\ eventrate
medical error
adverse events
sentinel event rate
event per people
Time to define
medical error % of adverse
events
% of sentinel event
sentinel events
Figure5. Subsystem2: dynamics of errors
Figure 5 shows a section of the stock and flow structure, subsystem 2, describing the dynamics of
errors in health care. The central stock “potential medical errors” change trough the inflow
“potential medical error rate” that is the result of the amount of treatments of the year “patient
treatment rate” and the “effect of patient safety index on medical error’. This last variable
represents the weight of the patient safety culture in the generation of clinical errors. It depends by
the stock “patient safety index”. We decided to create a graph function to explain the effect of the
patient safety index on medical errors. We assume that when the value of patient safety index is at
his maximum (100%) there are no medical errors, so the value of the “effect of patient safety index
on medical error” is zero, it means zero error each treatment. When the value of patient safety
index is at his minimum (0%) the value of the “effect of patient safety index on medical error’ is 1, it
means one error each treatment. However we decided to limit the value of the graph output (“effect
of patient safety index on medical error”) in a range from 0,003 and 0,8 because the two extreme
values do not represent the reality, we cannot have one error each treatment and furthermore we
cannot exclude the small chance that an error may still occur because hospitals and health care
are usually high-risk contexts, “errors can occur even in the best organizations because of the
fallibility of people” (Reason, 2000).
The value of the stock “patient safety index” was measured in the ward on 2013 and it was 66,52%
based on the scale means of the questionnaire’s answers. However we assumed that this value
was the result of the last 3 year CRM policies implement by the hospital, so the value of the stock
at the timeO (01/01/2011) of the simulation was set at 43%. This value arises from some info about
the CRM level before 2011collected during the interview with the CRM’s hospital key actors. As a
consequence of the last death occurred in the ward (24/12/2010), the last of nine deaths occurred
in the ward over the three years2008,2009 and 2010, the Obstetric and Gynecology ward's activity
were suspended (suspension did not affect emergencies cannot be postponed) for about a month,
during which the hospital’ management analyzed the “roots” causes of all the adverse events
occurred in the ward in 2010. The critical factors highlighted by the analysis of the root causes
have been the subject of planning by the Strategic Management. The Strategic Management, has
scheduled a number of organizational interventions both structural (the delivery room and
operating room were located on the same floor with the aim to optimize time, transfers and
interventions) and functional (increase in staff, review of procedures relating to hospital and pre-
hospitalization, improvements in staff availability). The launch of this innovative master plan in the
beginning of 2011allowed to achieve the following results: Reorganization of human resources
(mentoring activities through the establishment of a special team that has been focusing on the
evaluation of staff employed by the ward in terms of skills and experience; training with regular
meetings on specific CRM issues and the application of best practices; the improvement of the
organizational wellness and the organizational culture in term of patient safety), Quality and clinical
risk management (spontaneous reporting of near miss, no harm events and adverse events
through the Incident Reporting protocol; systematic review of medical records; activation of the
simes data stream; implementation of best practices, procedures, operational protocols, guidelines,
etc.), Integration of hospital / territory (for this purpose, has been defined Birth Path Protocol, which
provides the integrated management of pregnancy by: the management of low-risk pregnancy at
the Family services until the 36th week; the activation of an outpatient clinic for high-risk pregnancy
at the hospital; referrals to Family services for care of the mother and newborn after hospital
discharge).
From the stock “potential medical errors” there are 4 outflow “sentinel event rate”, adverse event
rate”, "near miss rate” and “no harm event rate” and the respective stock that represent the
different kind of error in which a hospital can occur. Each flow is regulated by the frequency
percentage of different errors in the medical routine (based on literature data analysis and the
historical data analysis of the hospital: 0,2% sentinel events, 5% adverse events, 78,8% near miss
and 16% no harm events).
Since our research scope is to test the weight of patient safety culture in the generation of clinical
errors, the model boundaries are set to create a simple but sensible SD model. For limitations of
the model see the section 7. “Conclusion, Implications, Limitations and further research”.
6.4 Scenario Analysis
Based on the analysis of the system described in the previous section, four different policies have
been tested and compared with the purpose to estimate the effect of such policies on hospital’s
performance. In effect, the amount of medical errors can be considered as a performance index for
the hospital, it tells us how the hospital personnel and how the hospital management deal with
patient safety. Indicators for performance and outcome measurement allow the quality of care and
services to be measured. This assessment can be done by creating quality indicators that describe
the performance including the occurring, for instance medical errors.
Scenario % of corrective action over % of Investment in
clinical event CRM policies
Base run scenario 10
No policy run scenario 0 0
Medium policies run scenario 20 (01/01/2011--31/12/2014) 10 (01/01/2011--31/12/2014)
40 (01/01/2015—01/01/2018) 16 (01/01/2015—01/01/2018)
High policies run scenario 20 (01/01/2011--31/12/2014) 10 (01/01/2011--31/12/2014)
100 (01/01/2015—01/01/2018) 20 (01/01/2015—01/01/2018)
Table 4. Scenario analysis
As showed in table4, these policies differ by the percentage of corrective action over clinical event
and by the percentage of investment in CRM policies. The percentage of corrective action over
clinical event refers to the number of corrective action effectively implemented after a reporting
(simes data stream or incident reporting), it ranges from 0 (no corrective action after reporting of
clinical error) to 100 (a corrective action each event reported). The percentage of investment in
CRM policies refers to the percentage of investment, respect the total amount of investment of the
hospital, reserved to CRM. It ranges from 0 (no investment in CRM policies) to 20 (20% of
investment in CRM policies). From the interrelation of these indexes four scenarios arise:
- In the “base run scenario”, the hospital policy is aimed at maintaining the current level of
CRM.
- In the “no policy scenario”, it is assumed that the hospital decides to cut to zero the actual
investment in CRM policies and as a consequence no corrective action will be implemented
over clinical event.
- Inthe “medium policies run scenario”, it is assumed that the hospital will increase the CRM
policies investment by 6% (from 10% to 16%) with the respect of the current level by the
beginning of 2015. The amount of corrective action implemented over clinical event will
increase by 20% with the respect of the current level by the beginning of 2015 (form 20% to
40% of corrective action implemented after an event reported).
- Inthe “high policies scenario”, ), it is assumed that the hospital will increase the CRM policy
investment by 10% with the respect of the current level, by the beginning of 2015 (from
10% to 20%), and the corrective action implemented over clinical event by 80% with the
respect of the current level, by the beginning of 2015 (from 20% to 100%, it means a
corrective action each single event reported).
For the scenario analysis a seven year time horizon is considered. The first three years of the
simulation run (2011 - 2013) have the scope to replicate the past ward’s performance. The last four
years (2014 — 2017), are intended to forecast the potential effect of chosen policies on ward
performance.
The following figures (fig. 6, 7 and 8) show the simulation results of the four different scenarios.
In the base run scenario, it is assumed that the hospital policy is aimed at maintaining the current
level of CRM. This policy could imply some minor error that determines an increase in hospital
attractiveness and in the patient safety index. Hospital attractiveness passed from 49,12% on 2011
to 55,60% on 2018, and patient safety index passed form 43% to 90%, as a result also the number
of patient per year increase. Despite the increase of patients, however, sentinel event passed from
0,44 event per year to 0,09 event per year and adverse event passed from 11 event per year to
2,34 event per year, it means an improvement of the patient safety and the quality of the service
delivered.
2011 2012 2013 2014 2015 2016 2017
2041 2012 2012 2014 2015 2016 2017
88 /yr
: 34
| 2
2 so
3 —
£ ae.
48 i citeere
aa 12 13 14 5 16 7
Figure 6. Base Run (Reference) & No CRM Policies Run (Current) for Hospital
In the no policy run scenario (figure6) we assume that starting from timeO (2011), the hospital
decide, to cut to zero the actual investment in CRM policies and as a consequence no corrective
action will be implemented over clinical event. This choice determine a gradual deterioration of
patient safety quality, the patient safety index drops to 30,30% by the end of 2017. As a
consequence the sentinel event increases till reaches 0,70 events per year and the adverse event
increases till reaches 17,64 event per year. Due of this bad situation the hospital attractiveness
drops to 47,59% by the end of 2017 resulting in a decline of patients per year. The analysis of this
scenario is aimed to test the model robustness under extreme conditions.
In the medium policies run scenario (figure7), it is assumed that the hospital will increase the CRM
policies investment by 6% (from 10% to 16%) with the respect of the current level by the beginning
of 2015. As a consequence the amount of corrective action implemented over clinical event will
increase by 20% with the respect of the current level by the beginning of 2015 (form 20% to 40% of
corrective action implemented after an event reported).
event/yr
10+
i
i i
i
2oia 2oiz 2013 2014 2015 2oie 2017
wanes
4 oat i
i ml a
_
o2+
oir <= —_
2012 2012 2018 2014 2015 zo1e 2017
he
°
G@ os
i S14
i i 3 is 5 ie 7
Figure 7. Base Run (Reference) & Medium CRM Policies Run (Current) for Hospital
As a result of adopting these policies a positive effect on all previously examined performance
indicators is registered. The patient safety index increases till 99,9% by the end of 2017 with a
decrease in sentinel event (0,06 event per year) and in adverse event (1,59 event per year). The
increase in hospital attractiveness caused by the low number of potential error generates an
increase in the patient inflow.
In the high policies run scenario (figure 8), it is assumed that the hospital will increase the CRM
policy investment by 10% with the respect of the current level, by the beginning of 2015 (from 10%
to 20%), and the corrective action implemented over clinical event by 80% with the respect of the
current level, by the beginning of 2015 (from 20% to 100%, it means a corrective action each
single event reported). As a result of this policy the patient safety index jumps to 100% by the end
of 2016 and stabilize it-self to this value, sentinel event and adverse event reduce themselves till
respectively 0, 05 and 1, 47 events per year. As a consequence hospital attractiveness increases
till 58,35%. However if we compare this scenario with the medium policies run scenario, there is no
a huge difference in term of performance, for this reason we can assume that the higher
investment cost required by this policy would not be counterbalanced by the low difference, in term
of performance, respect to the medium policies run scenario. Consequently, the management,
could prefer medium policies run scenario in term of sustainability.
2011 2012 2013 2014 2015 2016 2017
event/yr
0.4 rig,
- oe
0.2
0,2 ia
2014 2012 2013 2014 2015 2016 2017
eye
38
36
34
ae
i 50
nh 12 13 14 3s 16 7
Figure 8. Medium CRM Policies Run (Reference) & High CRM Policies Run (Current) for Hospital
6.5 Can an organizational behavior be better captured by a simulation results than
organizational historical data?
If we compare the behavior of the system during the first three years (2011, 2012, 2013) with the
historical data collected from the hospital, we can see that the amount of medical events arise from
the simulation is quite different from the number of events collected by the hospital (see table 5).
The model reproduces a number of adverse, no harm and near miss events that fit only partially
with the historical data collected from the hospital (tab. 5). More in detail, whereas we can observe
that the number of sentinel events during the 2011 — 2013 time range and the number of adverse
events of 2011 are the same, both in the simulation run and in the hospital historical data, the
number of adverse, no harm and near miss events showed by simulation model are much higher
than the ones collected from the hospital (tab. 5).
To explain this partially unfit of the model behaviors with the historical data, we can assume that:
a) the fit between simulation results and ward historical data (2011 — 2013 time horizon) about the
number of sentinel events could be ascribed to the national current regulation that oblige the
hospital staff to report a sentinel event, where the physicians and nurses always adhere to such a
regulation since sentinel events are very often together with patient’ claims;
b) the fit between simulation results and ward historical data (2011 time horizon) about the number
of adverse events (and the unfit for no harms and near miss events), could be ascribed to the
launch of a innovative master plan in the beginning of 2011, specifically set for the personnel of the
obstetric and gynecological ward, focused on the implementation of the incident reporting system
and SIMES data stream. The fact that only the number of adverse events showed by the
simulation model fits with the ward historical data could be explained exploring the ward’s staff
mental models about the role of the different typologies of events in generating medical errors,
where the adverse events are worth considering since they are the only ones that could generate
an observed medical error and, as consequence, a patient's claim.
So, we hypothesize that this partially data unfit is given by the high sensibility of the model to catch
the causal relationship between the organizational behaviors of the ward’s personnel and the
underlying mental models that generate them. The poor attention to CRM practices showed by the
ward's personnel can be confirmed also by the fact that during the questionnaire administration,
they sometimes argued about the inutility to submit an incident report especially when a no harm
event or a near miss have been occurred. As two ward’s nurses said during the questionnaire
administration: “At the end no one was hurt, so it doesn’t matter...”; “Even the patient doesn’t
realize that he was been in trouble, so why | have to report something about it?”.
Types of clinical adverse event Base run simulation results Hospital historical data
11 (2011) 10 (2011)
ADVERSE EVENTS 9 (2012) 0 (2012)
8 (2013) 1 (2013)
35 (2011) 3 (2011)
NO HARM EVENTS 31 (2012) 1 (2012)
26 (2013) 0 (2013)
173 (2011) 3 (2011)
NEAR MISS 154 (2012) 0 (2012)
126 (2013) 0 (2013)
0 (2011) 0 (2011)
SENTINELL EVENTS 0 (2012) 1 (2012)
1 (2013) 0 (2013)
Table 5. Comparison between base run simulation result and historical data of the first three years.
From the organizational psychology perspective, we can assume that the reason of this poor
attention to CRM practices lies on the resistance to change, which could be defined as the act of
opposing with modification or transformation of the organizational status quo. Resistance to
change can occur when people don’t really understand the reasons behind the change, when
people know the reasons for the change, but they don’t see how those reasons translate into
benefits they value, when people are unclear about how the change will impact their job roles, what
new expectations they will have to meet after the change and whether they will be able to meet
those expectations, when people feel that change plans are being imposed on them by others, that
they do not “own” the change.
This means that the first step in the organizational change process is the change of organizational
culture, a change that is not yet fully realized in the analyzed ward. Actually, the prevalent
organizational culture seems still to be the “blame culture”, where the error is seen as a failure and
not as something from which to learn.
Moreover, submit an incident report takes time, too much time for a nurse or a doctor, fully busy
with his work. The Incident Report form is very detailed and no ever simply to fill with the info
available, so often a staff member prefers to skip to submit a event without harm with the patient
and concentrate on his work.
7. Conclusions, Limitations, and further research
We suppose that our study could have a broad appeal for researchers. Globally the healthcare
cost increases due to the increase in aging population, total health care spending in the overall
developed countries is expected to increase dramatically (for example, in US it will reach $4.8
trillion in 2021, up from $2.6 trillion in 2010 and $75 billion in 1970. To put it in context, this means
that health care spending will account for nearly 20 percent of gross domestic product, or one-fifth
of the U.S. economy, by 2021). The complexity of the profit maximization phenomenon at the
expense of patient's safety has become a pressing issue requiring remedial action.
Our research findings suggest that it would be convenient for healthcare companies to invest in
CRM policies in order to reduce the expenses resulting from ineffective errors management
(claims, loss of image, loss of patient, insurance costs etc.).
Similar to other studies, our research has some limitations that can be addressed in the future
research work. One limitation is given by the model boundaries. In the model there is a lack of
financial indicators and some system variables that may influence the hospital’s performance have
not been considered (role of Regional Healthcare Administration, patient associations, J oint
Commission International). A second limitation is given by the choice to represent the different
patient safety index factors by a single variable that takes into account their average values.
Nevertheless, despite these limitations, this research delivers results with implications for the
applications of system dynamics methodology to CRM. It would be advisable to develop CRM
research at international level to undertake comparative research studies in cross cultural setting
aimed at creating communities of practice where sharing best result and fostering organizational
learning.
Acknowledgements: Our research team would like to express thanks and gratitude for all the assistance
and help provided by the following individuals:
Dr. Rocco Billone - Head Physician of the obstetric and gynecology ward,
Dr. Pietro Murgia - Associate Medical Director of the obstetric and gynecology ward,
Dr. Alessandro Aiello - Responsible of Midwife’s Nursing Care of the hospital, the whole staff of the obstetric
and gynecology ward of the hospital.
REFERENCES
Former Australian Council for Safety and Quality in Health Care (2003). "Open disclosure" Standards.
A national standard for open communication in public and private hospitals, following an adverse
event in health care. Commonwealth of Australia.
Albolino, S., Tartaglia, R., Bellandi, T., Amicosante, A., Bianchini, E., & Biggeri, A. (2010). Patient
safety and incident reporting: survey of Italian healthcare workers. BMJ Quality and Safety in
Health Care. , 19, 8-12.
Ceresia, F., & Montemaggiore, G. B. (2011). A system dynamics model for the CRM: a preliminary
analysis. The 29th International Conference of the System Dynamics Society. Washington, DC.
Cook, R., & Rasmussen, J . (2005). "Going Solid": a model of system dynamics and concequences of
patient safety. BMJ Quality and Safety in Health Care , 14, 130-134.
Greenberg, C., Regebogen, S., Studdert, D., Lipstz, S., Rogers, S., Zinner, M., etal. (2007). Patterns
of communication breakdowns resulting in injury to surgical patients. J ournal of American College
of Surgeons. , 204, 533-540.
Goldsmith D., Siegel M. (2011) Improving Health Care Management Through the Use of Dynamic
Simulation Modeling and Health Information Systems. System Dynamics conferences 2011
proceed papers.
Henriksen, K., Dayton, E., Keyes, M., Carayon, P., & Hughes, R. (2008). Understanding adverse
events: a human factors framework. In R. Hughes, Patient safety and quality: an evidence-based
handbook for nurses. Rockwille MD: Agency for Healthcare Research and Quality (US).
Homer, J ., & Hirsch, G. (2006). System dynamics modeling for Public Health: background and
opportunities. Americal J ournal of Public Health , 96 (3), 452-458.
Kohn, L. T., Corrigan, J ., & Donaldson, M. (2000). To err is human: building a safer health system.
Washington DC: National Accademy Press.
Leape LL. (1994). Error in medicine. J ournal of the American Medical Association, 272(23):1851-7
Nieva V.F., Sorra J . (2003) Safety culture assessment: a tool for improving patient safetyin healthcare
organizations. Qual Saf Health Care2003;12(S uppl I!):ii17—-ii23
Pellegrino, F. (2011). Risk management e comunicazione: cosa non va? Rivista della Societa Italiana
di Medicina Generale , 1, 19-20.
Reason, J ames (1990): Human Error. New York, NY, Cambridge University Press.
Reason, J . (2000). Human error: models and management. BMJ , 320, 768-770.
Reason, J . (2004). Beyond the organizational accident: the need for "error wisdom" on the frontline.
BMJ Quality and Safety in Health Care , 13, 28-33.
Stanhope, N., Vincent, C., Taylor-Adams, S., O'Connor, A., & Beard, R. (1997). Applying human
factors methods to clinical risk management in obstetrics. British J ournal of Obestrics and
Gynecology. , 104, 1225-1232.
Vennix, J ., Andersen, D., Richardson, G., & Rohrbaugh, J . (1992). Model-building for group decision
support issues and alternative in knowledge elicitation. european J ournal of Operational Research.
, 59 (1), 28-41.
Vincent, C. (2010). La sicurezza del paziente. Firenze: Springer.
Vincent, C. (2003). Understanding and responding to adverse events. New England J ournal of
Medicine , 348, 1051-1056.
Vincent, C., Neale, M., & Woloshynowych, M. (2001). Adverse events in British hospitals: preliminary
retrospective record review. BMJ , 322, 517-519.
Vincent, C., Taylor-Adams, S., & Stanhope, N. (1998). Framework for analysing risk and safety in
clinical medicine. BM) , 316, 1154-1157.
ANNEX 1 correlations
Organi Hazel
zational | Nonpun Supervisorim | Feedback equa Mariage Teamw | Hospita
Tabled im gu” |e. [Percept |expecotons | Communic | Communic Preyaf | ment | Eerocs | andor | Paint
gorretation | wor | S%9_ | Continu | RESP" | ons of ation SHO ase (fore | Ube | Hoswt [ea | BRR
k ous r Safety | promoting About Ps Pr Pauent H Transiti
improv. | Er safety Error ng stety_ | Units
ement
Team Work 7
--489" 0,222 | 0,015 0,055 0,078 -0,014 0,276 0,337 [0,239 | 0,241 0,014 | 0,295
0,01 0.265 |0,941 | 0,784 | 0,701 0,946 0,164 0,086} 0,23 0,227 |0,944 | 0,135
27_|27 21 27 27 27 21 27 21 27 27 7 27
Staffing
1 0085 |-0178 }0,279 | 0,135 -0,106 0,091 -0,156 |-0,145 | 0,022 |0,024 | 0,108
0,675 |0,375 0,159 | 0,502 0.6 0,651 0,436 [0,471 | 0,912 |0,907 | 0,592
27 21 27 27 27 27 27 21 27 27 27 27
Organizatio ' "
al 1 0,147 0,338 | 0,094 568 0,083 584” | 507 0,272 0,141 | 597
Learning—
Continuous 0,464 | 0,085 | 0,641 0,002 0,68 0,001 |0,007 | 0,17 |0,483 | 0,001
improvemen
t 21 27 27 27 21 27 21 27 21 27 21
Nonpunitive
Response 1 0,366 | 0,115 0,345 0,146 0,306 | 414 0,362 |0,104 | 0,249
To Error
0,061 | 0,567 0,078 0,467 0,121 | 0,032 | 0,064 | 0,605 | 0,21
21 27 27 21 27 21 27 27 7 27
Overall re F a j
Perceptions a 508” | ,549 0,262 ATs | 551 0,323 [0,347 | 444
of Safety
0,007 0,003 0,187 0,013 | 0,003 | 0,101 0,076 | 0,02
27 27 27 27 21 27 27 7 27
Supervisor " - a a
1 576 588" | 0,229 445] 558°} 482"] 0,351
expectation
§ & actions 0,002 0,001} 0,251 0,02] 0,002] 0,011] 0,072
promoting
safety 20 21 27 20 27 2 21 27
Feedback , 7 i .
snd 1 422°] 669 642"} 0,368] 386°] 524
Communica
tion About 0,028 0 0] 0,059] 0,047] 0,005
rror 27 27 27 27 21 21 27
Communica . 4 ,
1] 0,219 389°] 589" | 403] 0,147
Openness
0,274] 0,045] 0,001] 0,037] 0,463
27 21 21 21 27
Frequency .
of Event 1] zou" | 0,338] 0,321] 658
Reporting
0] 0,084] 0,103 0
27 27 21 21 27
Hospital . . .
Mangeniel 1 689" | 621" | 637
t Support for
Patient 0 0,001 0
Safety 27 21 27 27
Teamwork 7
Across 1] ,668"] 0,268
Hospital
Units o} 0,176
20 27 27
Hospital
Handoffs & 1 0,169
Transitions
0,399
21 27
Patient
Safety 1
Grade
21
